Search Console renamed the old Coverage report. The same job now lives under Indexing, in the Pages report, often still called the Page indexing report in Google's own docs. It sorts every URL Google knows about on the property into indexed and not indexed, and it attaches a reason to the URLs that are not indexed. The mistake is to treat every reason as a ticket. Several reasons are the report telling you that a rule you wrote is working.

Start with the not-indexed table, sorted by how many URLs sit in each reason. Open a reason, look at the examples, and only then inspect two or three of them. Filter by sitemap when you want the chart limited to URLs you submitted. A reason that is huge in the whole property and tiny inside your article sitemap is probably parameter URLs and tag pages, not the stories. The sitemap guide is how that filter becomes meaningful. The URL Inspection guide is how you read a single example.

After a real fix is on the template, use "Validate fix" on that reason. Validation rechecks a sample. It fails when another template still emits the old response. It is not sped up by requesting indexing on every example URL.

Server and access errors

These reasons mean Google tried to fetch a URL and the response was a failure or a refusal. On URLs you want indexed, they are repairs.

Reason What happened What to change
Server error (5xx) The host failed the request Logs, timeouts, capacity. 5xx also slows crawling.
Redirect error A loop, a chain that is too long, or a broken target One hop to a live URL. Google generally follows up to 10 hops.
Blocked due to unauthorized request (401) The URL wants a login Remove the login for public pages, or stop linking and listing private URLs.
Blocked due to access forbidden (403) The server refused the crawler Firewall, CDN, or bot rules that treat Googlebot as hostile.
Blocked due to other 4xx issue Some other client error Read the status on a sample. Google ignores the body of 4xx responses.

Do not use 401 or 403 to slow Google down. The status-code documentation says those codes, unlike 429, do not limit crawl rate. 429 is treated as overload and does slow crawling. If a public article returns 403 only to Googlebot, you have a bot rule, not an indexing mystery.

A redirect error on a redesigned site is often a chain: old path, intermediate path, campaign host, then the article. Point the old path at the final URL. Internal links should use that final URL too, so new crawls do not rebuild the chain. The internal linking guide covers link targets that still redirect.

Rules you may have meant

Reason What it means When to act
Excluded by 'noindex' tag The page asks not to be indexed Leave it if the page is a preview, a utility, or a true duplicate you do not want. Remove the tag if the page is an article.
Blocked by robots.txt Crawling is disallowed Leave it if you do not want the URL fetched. Allow it if you wanted noindex or a canonical to be seen.
Indexed, though blocked by robots.txt The URL is in Search without a crawl of its content Allow crawling and use noindex if the goal is removal.

A sudden wave of "Excluded by noindex" after a launch is a staging tag in the production template. A sudden wave of "Blocked by robots.txt" is the staging disallow. Both are file and template bugs, described in the robots.txt guide. Validate only after production HTML no longer contains the tag and the live robots.txt no longer matches the path.

Reasons that are usually healthy

Page with redirect. The URL redirects. Old addresses should do that. You care when the target is wrong or when the URL should have been the canonical itself.

Alternate page with proper canonical tag. The URL points at another URL as canonical, and Google accepted that. Check the target, not the alternate. If the target is indexed, this row is the system working. If you did not mean to canonical, the canonical guide is the correction.

Not found (404). The URL does not exist. Google drops 4xx URLs from the index over time and crawls them less often. Let genuine deletions stay. Act when the URL should exist, when an important page still links to it, or when the sitemap still lists it. A 404 on a URL you replaced should be a redirect to the replacement, not a redirect to the home page.

Duplicates and soft 404s

Duplicate without user-selected canonical. Google found copies and you did not name a preferred URL. Add a self-referential canonical on the copy you want, and point the others at it, or redirect them if they should not be separate URLs.

Duplicate, Google chose different canonical than user. You named a preferred URL and Google kept a different one. The other signals disagree: links, sitemap, redirects, or two pages that are not actually duplicates. Make those signals match. Inspect one example and read both canonical fields before you edit a hundred URLs.

Soft 404. The status was success and the body looked empty or like an error message. This row should shrink because the status code or the body changed, not because you disallowed the URLs. The soft 404 guide splits "gone" from "empty template". Validate this reason only after a live test on a sample shows either a real article or a real 404.

Discovered, and crawled, currently not indexed

Discovered – currently not indexed means Google knows the URL and has not indexed it. On a small site this often clears after the URL is linked from a page that is already crawled. On a large site, or a site whose URLs change daily, a large count here is one of the situations Google's crawl budget guide is written for. The thresholds in that guide are rough (very large sites, or medium sites that change quickly), not a line you must cross before you look. The crawl budget guide is about reducing URLs that should never have been discovered.

Crawled – currently not indexed means Google fetched the URL and did not keep it. Inspect a sample. If the sample is blocked, duplicated, or a soft 404, you are in an earlier section and the label is stale or mixed. If the sample is a clean article, the page is losing a quality or duplication judgment. Merge thin posts into a stronger URL, noindex or delete empty archives, and stop publishing the same brief update at five URLs. The indexing diagnostic is the per-article version of that choice. The report is how you see it is happening to a whole directory.

An order that avoids busywork

  1. Fix 5xx, 401, 403, and redirect errors on URLs that should be public documents.
  2. Remove accidental noindex and accidental robots.txt blocks. Keep the ones you meant.
  3. Repair canonical conflicts on the templates that own the duplicate reasons.
  4. Turn soft 404s into real 200 documents or real 4xx responses, and take the dead URLs out of the sitemap.
  5. For "crawled, not indexed" samples that are technically fine, edit or merge the content.
  6. Leave healthy redirects, proper alternates, and genuine 404s alone.

Export examples and group them by path prefix before you assign the work. /tag/ and /news/ can share a reason and need different owners. Write down which template you changed so the next redesign does not restore the tag.

When validation fails, open the failing examples. They are often a second host, a second language, or an AMP or print template that nobody updated. Fix that template and validate again. Requesting indexing on the failures without a template change recreates the same row.

The Page indexing report is a sort of the URLs Google already knows. It is finished when the reasons left on your article sitemap are the ones you would explain to a colleague without apologizing: old redirects, deliberate noindex, and the occasional page that was crawled and still needs a better article.

CrawlFaster groups indexing status across the URLs you track, which is useful when a single reason starts climbing inside one section. The report in Search Console remains the source of the reason names, and the fix remains on the template those URLs share.